Abstract
Background: Early studies of pediatric acute liver failure (PALF) estimate that 60%–70% of patients survive with their native liver (SNL), 20%–25% require an emergent liver transplant (LT), and 10%–15% are mortalities. These studies were largely limited to academic centers and focused on short-term outcomes.
Methods: Herein, we analyzed California Department of Healthcare Access and Information data, which is demographically representative of the US population. Between June 1990 and December 2018, 2270 children were admitted to California hospitals with PALF.
Results: The overall incidence of PALF was 16.75 per 100,000 live births, with no evidence of seasonality. LT was performed in 188 (8.3%) patients and 425 (18.7%) patients died. Mortality was associated with contraindications to transplant, such as concomitant hematologic malignancy or cardiac disease. With a median follow-up of 4.5 years, nearly all deaths occurred within 4 months of presentation, with no long-term liver-related comorbidities in patients with SNL. Implementation of pediatric end-stage liver disease/MELD-based organ allocation in 2002 resulted in a more than 2-fold increase in the likelihood of LT in children <10 years (subdistribution hazard ratio: 3.35). Patients receiving care at transplant centers after this change were younger (10 vs. 12 y), more often from urban areas (98.2% vs. 94%), and more likely to receive LT for metabolic, genetic, or “other” causes of PALF (up to subdistribution hazard ratio 6.4 [1.84–22.26], p<0.01). While mortality was higher at transplant centers (26.5% vs. 14.6%), the time to mortality was also increased (31 vs. 7 d), likely reflecting disease complexity requiring management in a specialized children’s hospital.
Conclusion: In conclusion, this study supports an estimate of incidence and long-term outcomes of PALF across a large population and highlights the role of specialized pediatric transplant centers for the optimized care of PALF.
Keywords: epidemiology, PALF, pediatric liver transplantation, posttransplant outcomes
INTRODUCTION
Pediatric acute liver failure (PALF) is a rapidly progressive disease that affects children of all ages and can lead to multisystem organ failure and death. While most children will recover, initial estimates indicate that up to 25% will require emergent liver transplant (LT) and at least 10% will die.1 Although treatment is typically supportive, LT remains the definitive cure for PALF, representing 10%–15% of all pediatric LTs performed each year in the United States.2,3,4,5 Initial screening for PALF is typically performed in accordance with NIDDK-funded Pediatric Acute Liver Failure Study Group (PALFSG) criteria using clinical and laboratory findings (elevated liver enzymes and coagulopathy), followed by adjunctive testing to narrow down the specific etiology.6,7 Recognized causes of PALF include viral infections, genetic disorders, drug toxicity, autoimmune hepatitis, metabolic disease, oncologic processes, and ischemia; despite a wide range of known causes, up to half of PALF cases are “indeterminate” and have been associated with worse outcomes.7,8,9
There have been several population studies within the PALF space, namely those that stem from North American and European cohorts.10,11 These studies establish multiple risk factors for LT or death, including neonatal age and select diagnostic bloodwork (such as international normalized ratio, bilirubin, and ammonia).11 These labs have formed the basis of ongoing efforts to further improve upon universal severity scoring (ie, MELD/pediatric end-stage liver disease [PELD]) to develop predictive algorithms for disease outcomes to drive treatment decisions.12,13,14 Along with improving diagnostic pipelines, mortality and LT outcomes across time in the PALFSG cohort and data from the Scientific Registry of Transplant Recipients (SRTR) have noted improved overall and waitlist mortality in the interim.2,10,15 These temporal changes are inclusive of strategic changes to organ allocation policies in the pediatric population, spanning the range of modifications to pediatric priority and organ allocation, implementation of severity scoring, and elimination of ICU admission as a requirement for Status 1 listing that have subsequently increased LT rates in PALF.15
Although short-term survival outcomes in the PALF population have improved in recent years, incidence and long-term sequelae in the North American population, especially outside academic centers, remain understudied. This study leverages comprehensive registry data maintained by the state of California, capturing more than 10% of the US population, to form a large cohort that tracks disease outcomes across demographics and over time.16 In doing so, this study aims to better characterize the incidence of PALF and establish clinical factors that may prognosticate patient outcomes.
METHODS
This study was approved by the California Committee for the Protection of Human Subjects (Project CS0002063) in accordance with both the Declarations of Helsinki and Istanbul. Data were obtained from the California Department of Healthcare Access and Information (HCAI) database, which is maintained by the state of California to improve health care quality.17 Data are required from all hospitals licensed in California and must have a data error rate of <2%. Data error rates are ensured at the time of submission at the state level, and identified errors under the 2% limit from compliant institutions may be subject to deletion.18 The HCAI data set encompasses claims from patient discharge, emergency department, ambulatory surgery data, and linked death records.
Database characteristics with inclusion and exclusion criteria
Data were obtained from Patient Discharge, Emergency Department, and Ambulatory Surgery data sets, which include patient records ranging from June 1990 to December 2018 for Patient Discharge and from January 2005 to December 2018 for Emergency Department and Ambulatory Surgery. International Classification of Diseases (ICD) Codes, 9th and 10th Revisions, were used to identify acute liver failure cases and included the following: 570, K7200, K7201, and K720. Inclusion criteria consisted of pediatric patients <18 years of age at the time of hospital admission with PALF diagnostic codes. Exclusion criteria consisted of patients ≥18 years at the time of presentation. Quality control was established to exclude cases with LT before PALF diagnosis, adjacent but nonrelated diseases like biliary atresia, or erroneous records, such as patients with a procedure date occurring after the date of death. A CONSORT diagram is outlined in Figure 1.
FIGURE 1.
CONSORT diagram of patients with PALF included for analysis from the California HCAI database from 1990 to 2018. Flowchart outlining identified patients from the California HCAI database from 1990 to 2018, including excluded populations. A total of 2270 patients were included in the final analysis. Abbreviations: HCAI, Health Care Access and Information; PALF, pediatric acute liver failure.
Clinical variables and outcomes
Demographics, clinical comorbidities, county of residence, age at time of illness, etiology, and discharge diagnoses were examined. Urban or rural residence was derived from the patient’s county of residence using classifications from the California State Association of Counties.19 Diagnosis codes were grouped into the following categories: viral hepatitis, acetaminophen toxicity, other immune-related (autoimmune hepatitis and hemophagocytic lymphohistiocytosis), metabolic, and other (neoplasm, Budd-Chiari, hepatic infarction, sepsis, shock, and other drugs and toxins). Cases with unidentified etiology were classified as “indeterminate.” LT and mortality after PALF diagnosis were the primary study endpoints. In examining patient survival after PALF diagnosis, patients were censored at transplant. Median time to death after PALF was 0.05 years [IQR: 0.01–0.34] or ~18 days. Based on the overall distribution of death after PALF admission, we chose the third quartile, or 4 months, as the cutoff in defining long-term survival, and the causes of death among long-term survivors were assessed. ICD10 codes for causes of death included the following: Hematological malignancy: C859, C910, C920, C959, and D469; cardiac: I420, I429, I500, Q213, Q230, Q234, and Q249; gastrointestinal complication: K631, K746, K766, K781, K839, K902, and Q793; solid tumor malignancy: C240, C349, C419, C499, C719, C749, C80, and C97.
Statistical analysis
Continuous measures were reported with mean and SD or, in the case of skewed distributions, with median and IQR. For categorical measures, counts and percentages were displayed. Per HCAI data use guidelines, variables with <11 patients were reported as “<11” rather than the discrete value; however, statistical analyses on these variables were performed using the actual value. PALF incidence rates were calculated using total live births from hospital annual utilization reports. Median follow-up from PALF diagnosis was calculated using the reverse Kaplan-Meier method. Multivariable regression analyses were performed using Fine-Gray subdistribution hazard models to estimate cumulative incidence with LT and death as competing events.20 The patient cohort was further subsampled to assess 3 additional parameters: the introduction of the PELD and MELD scoring systems in 2002, Status 1A/1B policy change for pediatric LT, including the elimination of ICU admission for Status 1A designation,21 and care at specialized transplant or nontransplant centers. The association between covariates and survival outcomes was analyzed in a stratified analysis, and comparisons were made within each comparative group. Analysis was conducted using R 4.0.2. p values ≤0.05 were considered statistically significant.
RESULTS
PALF incidence, seasonality, and risk of late liver-related complications
Across the 28-year study period, 2270 children were admitted to California hospitals with diagnostic codes consistent with PALF (Figure 1). The overall incidence of PALF was 16.75 per 100,000 live births based on birth data from 2012 to 2018 and comparable to previous estimates of 500–600 cases per year in the United States.6 Dates of admission were spread equally between the 4 seasons with no significant differences (p=0.56). 49.8% of patients were ≥10 years old at PALF admission and 49.3% were female. A minority of patients were non-Hispanic White (28.6%), with significant representation of Hispanic (44.3%), Black (10.4%), and Asian/Pacific Islander (7.4%) ethnic groups (Table 1). The most prominent etiology of PALF was indeterminate (55.4%), followed by “other” (27.3%), viral (7.8%), acetaminophen-induced (6.1%), other immune-related (2.2%), and metabolic (1.2%), with the acetaminophen and other immune-related proportions in our cohort lower than other estimates.6 Among 1692 patients whose county of residence was available, 95.2% lived in an urban environment. Median follow-up time after PALF diagnosis was 4.5 [IQR: 0.06, 10.4] years. Over the study period, 188 LT (8.3%) and 425 (18.7%) deaths were recorded. Of the 425 patients who died, 13 had received LT and 108 patients (5%) were <3 months of age at the time of presentation. The cumulative incidence of death and LT were highest within the first 6 months from initial PALF presentation (Figure 2). At 4 months, the cumulative incidence of LT was 6.3% and death was 17.6%. The longest follow-up from PALF diagnosis was quantified at 26.6 years (Figure 3). Patients who experienced survival with native liver (SNL) (survived 60 d without transplant) had a median follow-up of 2.3 [IQR: 0, 8.7] years compared to LT at 9.0 [IQR: 3.4, 13.8] years.
TABLE 1.
Demographics of the total California HCAI PALF study population from 1990 to 2018
| Demographics | Total population (N=2270) |
|---|---|
| Age at ALF (y), mean (SD) | 9.2 (6.4) |
| Distribution, y, N (%) | |
| <2 | 502 (22.1) |
| 2 to<5 | 303 (13.3) |
| 5 to<10 | 335 (14.8) |
| ≥10 | 1130 (49.8) |
| Female, N (%) | 1118 (49.3) |
| Race/ethnicity, N (%) | |
| White | 649 (28.6) |
| Black | 235 (10.4) |
| Hispanic | 1005 (44.3) |
| Asian/Pacific Islander | 168 (7.4) |
| Native American/Alaska Native | 90 (4.0) |
| Other/Unknown | 123 (5.4) |
| Etiology | |
| Indeterminate | 1258 (55.4) |
| Other | 619 (27.3) |
| Viral hepatitis | 178 (7.8) |
| Acetaminophen | 138 (6.1) |
| Other immune-related | 49 (2.2) |
| Metabolic | 28 (1.2) |
| Urbanizationa | |
| Urban | 1611 (95.2) |
| Rural | 81 (4.8) |
| Follow-up time, median in years [IQR] | 4.5 [0.06, 10.4] |
| Transplants, N (%) | 188 (8.3) |
| Deaths, N (%) | 425 (18.7) |
Note: The “Other” etiology category encompasses neoplasm, Budd-Chiari, hepatic infarction, sepsis, shock, and non-acetaminophen drug toxicity. “Other immune-related” includes HLH and autoimmune hepatitis. Standard deviation is indicated as “SD.”
Urbanization data were available for a subset of 1692 patients.
Abbreviations: ALF, acute liver failure; HCAI, Health Care Access and Information; HLH, hemophagocytic lymphohistiocytosis; PALF, pediatric acute liver failure.
FIGURE 2.
Cumulative incidence of liver transplant and death in the PALF population. A reverse Kaplan-Meier curve illustrating the overall incidence of liver transplant and deaths within the overall study cohort of 2270 patients in the HCAI PALF cohort. Both primary endpoint incidences reached a plateau within 6 months of diagnosis (as indicated by the vertical dashed line), with a total of 18.7% mortality and 8.3% receiving LT over the study period. Abbreviations: HCAI, Health Care Access and Information; PALF, pediatric acute liver failure.
FIGURE 3.
Long-term follow-up and mortality in the California HCAI PALF population. A Kaplan-Meier curve depicting long-term mortality in the total PALF cohort (N=2238). The longest follow-up to PALF diagnosis was 26.6 years. Abbreviations: HCAI, Health Care Access and Information; PALF, pediatric acute liver failure.
At the time of PALF diagnosis, 29 patients had underlying cardiac comorbidities, of which none received a LT or died; 40 patients had gastrointestinal comorbidities, with 16 receiving LT and no deaths. Late, non–PALF-related mortality was defined as death occurring >4 months after presentation and occurred in 104 patients. Late causes of death following PALF were largely attributed to hematologic malignancy (n=25) and cardiac complication (n=13) (Supplemental Table S1, http://links.lww.com/HC9/C355). Patients who achieved SNL did not appear to have evidence of late liver-related mortality in our cohort.
Risk factors for mortality and transplant in the PALF population
Multivariable analyses were conducted to assess covariate risk factors associated with mortality and transplant risk. Patients were additionally substratified according to pre- and post-2002 following the United Network for Organ Sharing (UNOS) policy for organ allocation using the MELD/PELD score to improve US congressional mandates for objective waitlist ranking. Before this, patients were limited to their locoregional organ procurement organization and were allocated based on admission/ICU status and waitlist time.15 While there was no difference in those who were <5 years of age at presentation, children 5–10 years of age (14.8% of patients) had a lower risk of mortality when compared to the reference cohort of children >10 years (subdistribution hazard ratio [sHR]: 0.72 [0.54–0.96], p=0.02) (Table 2). This effect was associated with the implementation of PELD/MELD score-based allocation (sHR: 0.66 [0.44–0.98], p=0.04) and higher rates of transplant for children <10 years of age (up to sHR 2.17 [1.39–3.41], p<0.001) (Table 3). Overall, female patients had a lower risk of mortality when compared to male patients (sHR: 0.80 [95% CI: 0.64–0.99], p=0.04, Table 2) with no difference in risk of transplant (Table 3). When compared to indeterminate PALF, patients with acetaminophen-induced liver failure had a lower risk of mortality (sHR: 0.43 [0.19–0.98], p=0.043, Table 2). In addition to having the highest mortality risk (Table 2), “other” causes of PALF were less likely to undergo LT (sHR: 0.51 [0.32–0.8], p<0.01). This is likely related to overall transplant candidacy as this category includes neoplasms, Budd-Chiari syndrome, hepatic infarction, sepsis, shock, and other drugs or related toxins (Table 3); these comprise the major categories of transplant ineligibility, including clinical instability (eg, uncontrolled sepsis), irreversible end organ damage (eg, terminal encephalopathy), and terminal extrahepatic disease (eg, multisystem mitochondrial disease).22,23
TABLE 2.
PALF mortality risk before and following PELD/MELD score implementation in 2002
| Total population (N=2270) | p | Pre-PELD/MELD: 1990–2002 (N=695) | p | PELD/MELD: 2002–2018 (N=1575) | p | |
|---|---|---|---|---|---|---|
| Age at ALF, y, sHRa [95% CI] | ||||||
| <2 | 1.14 [0.90–1.44] | 0.28 | 1.24 [0.69–2.24] | 0.47 | 1.18 [0.92–1.52] | 0.20 |
| 2 to <5 | 1.04 [0.79–1.38] | 0.78 | 0.71 [0.34–1.49] | 0.36 | 1.20 [0.83–1.73] | 0.34 |
| 5 to <10 | 0.72 [0.54–0.96] | 0.023 | 0.96 [0.46–2.02] | 0.92 | 0.66 [0.44–0.98] | 0.041 |
| ≥10 | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Sex, sHRa [95% CI] | ||||||
| Male | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Female | 0.80 [0.64–0.99] | 0.040 | 0.70 [0.42–1.14] | 0.15 | 0.83 [0.66–1.05] | 0.12 |
| Race/ethnicity, sHRa [95% CI] | ||||||
| White | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Black | 0.83 [0.57–1.20] | 0.32 | 0.70 [0.29–1.67] | 0.42 | 0.86 [0.56–1.32] | 0.49 |
| Hispanic | 0.87 [0.72–1.06] | 0.16 | 0.84 [0.49–1.43] | 0.52 | 0.85 [0.63–1.15] | 0.29 |
| Asian/Pacific Islander | 1.06 [0.72–1.54] | 0.78 | 0.92 [0.37–2.28] | 0.86 | 1.09 [0.69–1.73] | 0.72 |
| Native American/ Alaska Native/Other | 0.80 [0.52–1.23] | 0.31 | 0.51 [0.21–1.24] | 0.14 | 1.02 [0.63–1.66] | 0.94 |
| Etiology, sHRa [95% CI] | ||||||
| Indeterminate | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Viral | 1.32 [0.88–1.96] | 0.18 | 0.58 [0.07–4.65] | 0.60 | 1.63 [0.98–2.72] | 0.06 |
| Acetaminophen toxicity | 0.43 [0.19–0.98] | 0.043 | 0.19 [0.00–12.65] | 0.44 | 0.52 [0.23–1.17] | 0.11 |
| Other immune-related | 1.48 [0.68–3.2] | 0.33 | NA | NA | 1.48 [0.64–3.46] | 0.36 |
| Metabolic | 1.89 [0.85–4.18] | 0.12 | NA | NA | 2.55 [1.17–5.54] | 0.018 |
| Other | 3.66 [2.92–4.59] | <0.001 | 2.47 [1.49–4.09] | <0.001 | 3.87 [3.02–4.96] | <0.001 |
Italics denotes statistically significance (p<0.05).
Subdistribution hazard ratio (sHR). Analysis using Cox proportional hazards models demonstrates that age 5–10 years at presentation and female sex were associated with a decreased risk of mortality. Race and ethnicity were not associated with mortality. Acetaminophen toxicity had the lowest risk of mortality, while patients with “other” etiologies experienced the greatest mortality.
Abbreviations: ALF, acute liver failure; NA, not available; PALF, pediatric acute liver failure; PELD, pediatric end-stage liver disease.
TABLE 3.
PALF transplant risk before and following PELD score implementation in 2002
| Total population (N=2270) | p | Pre-PELD/MELD 1990–2002 (N=695) | p | PELD/MELD: 2002–2018 (N=1575) | p | |
|---|---|---|---|---|---|---|
| Age at ALF, y, sHRa [95% CI] | ||||||
| <2 | 2.17 [1.39–3.41] | <0.001 | 1.22 [0.61–2.43] | 0.58 | 2.76 [1.59–4.80] | <0.001 |
| 2 to <5 | 2.00 [1.35–2.94] | <0.001 | 0.93 [0.45–1.92] | 0.84 | 2.92 [1.57–5.45] | <0.001 |
| 5 to <10 | 2.04 [1.34–3.09] | <0.001 | 0.74 [0.33–1.70] | 0.48 | 3.35 [1.87–5.99] | <0.001 |
| ≥10 | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Sex, sHRa [95% CI] | ||||||
| Male | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Female | 1.23 [0.92–1.65] | 0.16 | 1.60 [0.97–2.64] | 0.07 | 1.01 [0.64–1.58] | 0.98 |
| Race/ethnicity, sHRa [95% CI] | ||||||
| White | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Black | 0.71 [0.41–1.23] | 0.22 | 1.02 [0.51–2.04] | 0.96 | 0.40 [0.10–1.68] | 0.21 |
| Hispanic | 0.69 [0.48–0.98] | 0.04 | 0.68 [0.37–1.26] | 0.22 | 0.75 [0.48–1.18] | 0.22 |
| Asian/Pacific Islander | 0.84 [0.44–1.61] | 0.60 | 1.05 [0.41–2.68] | 0.93 | 0.58 [0.24–1.42] | 0.23 |
| Native American/ Alaska Native/Other | 0.81 [0.48–1.37] | 0.43 | 0.62 [0.10–3.72] | 0.60 | 1.26 [0.50–3.19] | 0.62 |
| Etiology, sHRa [95% CI] | ||||||
| Indeterminate | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Viral | 2.75 [1.79–4.22] | <0.001 | 3.07 [1.56–6.03] | 0.001 | 2.51 [1.44–4.40] | 0.001 |
| Acetaminophen toxicity | 0.37 [0.13–1.05] | 0.06 | 0.45 [0.02–11.78] | 0.63 | 0.20 [0.00–48.85] | 0.57 |
| Other immune-related | 2.38 [1.11–5.11] | 0.026 | NA | NA | 3.07 [1.31–7.18] | 0.01 |
| Metabolic | 2.48 [0.94–6.53] | 0.07 | NA | NA | NA | NA |
| Other | 0.51 [0.32–0.80] | 0.003 | 0.52 [0.20–1.31] | 0.16 | 0.54 [0.31–0.94] | 0.031 |
Italics denotes statistically significance (p<0.05).
Subdistribution hazard ratio (sHR). Analysis using Cox proportional hazards models demonstrates that patients under 10 years of age at presentation were associated with a greater risk of transplant. Female sex before PELD, Hispanic race in aggregate, and Other Immune-Related cause were associated with a higher risk of transplant. “Other” cause of PALF was associated with a lower risk of transplant.
Abbreviations: ALF, acute liver failure; NA, not available; PALF, pediatric acute liver failure; PELD, pediatric end-stage liver disease.
Risk factors for transplant following UNOS Status 1A policy change in 2012
In addition to MELD/PELD scoring, several policy changes were implemented from 2005 to 2012 by UNOS to improve pediatric liver allocation on the basis of disease severity. Changes in 2005 split the highest urgency category (Status 1) into Status 1A and 1B, which prioritized acute liver failure as 1A over other transplant-eligible conditions, such as hepatoblastoma, chronic liver disease, and other metabolic defects grouped into a 1B category.15,21 This was further iterated in 2012 with a change that repealed the requirement for intensive care unit hospitalization for Status 1 listing.15,24 To examine whether these targeted policy changes affected transplant risk in PALF, a subset of patients was divided based on the time of diagnosis from June 2000 to February 2012 (n=375) and February 2012–2018 (n=599). No differences were observed among sex, race/ethnicity, urbanization, or number of performed transplants between the 2 eras (Supplemental Table S2, http://links.lww.com/HC9/C355). Multivariable transplant risk assessment revealed that patients 2–5 years of age were less likely to undergo LT before 2012 (sHR: 0.24 [0.06–0.95], p=0.04), which was resolved following the policy change (sHR: 1.97 [0.84–4.67], p=0.12) (Table 4). These results are similarly reflected in a US transplant registry–based analysis studying these policy changes, noting the greatest increase in pediatric LT in PALF associated with MELD/PELD as compared to the 2005–2012 changes.15
TABLE 4.
PALF transplant risk of a subset PALF before and following the Status 1A/1B listing policy change, including elimination of ICU admission for Status 1A eligibility in 2012
| Risk of transplant | Overall (N=974) | p | Status 1A/1B change 2000–2012 (N=375) | p | ICU req lifted 2012–2018 (N=599) | p |
|---|---|---|---|---|---|---|
| Age at ALF, y, sHRa [95% CI] | ||||||
| <2 | 1.00 [0.51–1.98] | 0.98 | 1.09 [0.4–2.98] | 0.87 | 1.04 [0.4–2.72] | 0.94 |
| 2 to <5 | 0.81 [0.39–1.68] | 0.57 | 0.24 [0.06–0.95] | 0.04 | 1.97 [0.84–4.67] | 0.12 |
| 5 to <10 | 0.58 [0.29–1.2] | 0.15 | 0.34 [0.08–1.38] | 0.13 | 0.77 [0.32–1.87] | 0.56 |
| ≥10 | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Sex, sHRa [95% CI] | ||||||
| Female | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Male | 1.38 [0.84–2.26] | 0.2 | 1.95 [0.84–4.52] | 0.12 | 1.07 [0.57–2.01] | 0.83 |
| Race/ethnicity, sHRa [95% CI] | ||||||
| White | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Black | 0.87 [0.3–2.6] | 0.81 | 1.38 [0.31–6.08] | 0.67 | 0.54 [0.11–2.66] | 0.45 |
| Hispanic | 1.34 [0.69–2.58] | 0.39 | 1.84 [0.65–5.23] | 0.25 | 1.21 [0.51–2.84] | 0.66 |
| Asian/Pacific Islander | 1.82 [0.72–4.59] | 0.2 | 0.58 [0.05–6.51] | 0.66 | 1.67 [0.56–4.91] | 0.36 |
| Native American/ Alaska Native/Other | 1.41 [0.18–0.95] | 0.75 | NA | 1.05 [0.12–9.14] | 0.97 | |
| Etiology, sHRa [95% CI] | ||||||
| Indeterminate | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Viral | 0.13 [0.02–0.95] | 0.04 | 0.55 [0.07–4.2] | 0.56 | NA | |
| Acetaminophen toxicity | 0.59 [0.08–4.42] | 0.61 | NA | 1.04 [0.13–8.05] | 0.97 | |
| Other immune-related | 1.22 [0.28–5.25] | 0.79 | 4.4 [0.84–23.15] | 0.08 | NA | |
| Metabolic/genetic | 3.05 [1.06–8.73] | 0.04 | 8.47 [1.84–38.93] | 0.006 | 2.04 [0.46–9.00] | 0.35 |
| Other | 0.13 [0.02–0.94] | 0.04 | NA | 0.14 [0.02–1.04] | 0.05 | |
Italics denotes statistically significance (p<0.05).
Subdistribution hazard ratio (sHR). Analysis using Cox proportional hazards models demonstrates that patients 2–5 years old before 2012 had a lower risk of transplant. Metabolic/genetic causes of PALF were associated with an increased risk of transplant, whereas other causes of PALF were associated with decreased transplant risk.
Abbreviations: ALF, acute liver failure; ICU, intensive care unit; NA, not available; PALF, pediatric acute liver failure.
Risk of mortality and LT at specialized transplant or nontransplant centers
We hypothesized that initial presentation at a nontransplant center may result in delays or lack of specialist care, thereby impacting outcomes. Thus, a second subgroup analysis was conducted stratifying the type of hospital patients were admitted to (transplant vs. nontransplant center) at the time of diagnosis. This analysis was conducted on patients diagnosed with PALF between 2009 and 2018 with available hospital admissions data. Among the 974 patients identified, patients with PALF initially presenting to transplant centers were younger, with a median age of 10 years [IQR: 3,14], than those at nontransplant centers (median age 12 years [IQR: 5,16] p<0.001) (Supplemental Table S3, http://links.lww.com/HC9/C355). No differences were observed with respect to the baseline characteristics of sex, race/ethnicity, and etiology. Patients admitted to transplant centers resided more frequently in urban areas (98.2% vs. 94%) (p=0.002) and were more likely to undergo LT compared to patients admitted to nontransplant centers (11.7% vs. 3.2% LT performed, respectively, p<0.001) (Supplemental Table S3, http://links.lww.com/HC9/C355).
Multivariable mortality risk was assessed for the initial admission center, and the risk of death was assessed. In this subgroup analysis, patients under 2 years of age in aggregate show an increased risk of mortality (sHR: 1.73 [1.03–2.91], p=0.04), regardless of type of center (Table 5). A greater mortality rate of 26.5% was associated with patients admitted to transplant centers as compared to nontransplant centers (14.6%; p=0.002), which is also associated with a longer interval from PALF diagnosis to death (31 d [IQR: 7,118]) versus nontransplant centers (7 d [IQR: 2,33]) (p=0.01) (Table 5). Analysis of risk for undergoing transplant revealed that children with metabolic and genetic etiologies of PALF were 6 times more likely to undergo transplant when initially admitted to transplant centers (sHR: 6.4 [1.84-22.26], p=0.003) (Table 6).
TABLE 5.
PALF mortality risk of a subset PALF population admitted to nontransplant or transplant centers from 2009 to 2013
| Overall (N=481)a | p | Nontransplant center (N=247) | p | Transplant center (N=234) | p | |
|---|---|---|---|---|---|---|
| Age at ALF, y, sHRb [95% CI] | ||||||
| <2 | 1.73 [1.03–2.91] | 0.04 | 1.46 [0.55–3.88] | 0.44 | 1.61 [0.84–3.10] | 0.15 |
| 2 to <5 | 0.64 [0.32–1.27] | 0.20 | 0.31 [0.07–1.36] | 0.12 | 0.75 [0.34–1.67] | 0.49 |
| 5 to <10 | 0.67 [0.38–1.19] | 0.17 | 0.72 [0.30–1.74] | 0.46 | 0.57 [0.27–1.23] | 0.15 |
| ≥10 | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Sex, sHRb [95% CI] | ||||||
| Female | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Male | 1.11 [0.74–1.67] | 0.60 | 1.28 [0.64–2.55] | 0.48 | 0.94 [0.55–1.60] | 0.83 |
| Race/ethnicity, sHRb [95% CI] | ||||||
| White | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Black | 1.44 [0.63–3.30] | 0.39 | 0.64 [0.13–3.13] | 0.58 | 2.55 [0.90–7.20] | 0.08 |
| Hispanic | 1.44 [0.81–2.57] | 0.22 | 1.08 [0.44–2.61] | 0.87 | 1.93 [0.88–4.24] | 0.10 |
| Asian/Pacific Islander | 4.25 [2.16–8.35] | <0.001 | 3.15 [1.15–8.65] | 0.03 | 5.94 [2.33–15.11] | <0.001 |
| Native American/Alaska Native/Other | 2.89 [0.65–12.83] | 0.16 | 1.93 [0.23–16.33] | 0.55 | 4.50 [0.54–37.35] | 0.16 |
| Etiology, sHRb [95% CI] | ||||||
| Indeterminate | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Viral | 0.71 [0.34–1.49] | 0.37 | 1.48 [0.50–4.33] | 0.48 | 0.36 [0.13–1.05] | 0.06 |
| Acetaminophen toxicity | NA | NA | NA | |||
| Other immune-related | 0.49 [0.12–2.07] | 0.33 | 2.60 [0.30–22.12] | 0.38 | 0.22 [0.03–1.63] | 0.14 |
| Metabolic/genetic | NA | NA | NA | |||
| Other | 1.62 [0.93–2.83] | 0.09 | 2.22 [0.97–5.11] | 0.06 | 1.40 [0.64–3.07] | 0.40 |
| Proportional comparison between nontransplant and transplant centers | ||||||
| Transplants, N (%) | 36 (7.5) | <11 | >11 | <0.001 | ||
| Deaths, N (%) | 98 (20.4) | 36 (14.6) | 62 (26.5) | 0.002 | ||
| Time ALF – death (d), median [IQR] | 20 [4, 78] | 7 [2, 33] | 31 [7, 118] | 0.01 | ||
Italics denotes statistically significance (p<0.05).
This set of patients is a sample (n=481, from 2009 to 2013) of the whole 974 patients that span the nontransplant center to transplant center comparison (from 2009 to 2018). Patients <2 years old and Asian/Pacific Islanders had a higher risk of mortality. Transplant rates were higher with increased mortality rate and duration between diagnosis and death in transplant centers.
Subdistribution hazard ratio (sHR).
Abbreviations: ALF, acute liver failure; NA, not available; PALF, pediatric acute liver failure.
TABLE 6.
PALF transplant risk of a subset PALF population from 2009 to 2018, admitted to nontransplant or transplant centers
| Overall (N=974) | p | Nontransplant center (N=564) | p | Transplant center (N=410) | p | |
|---|---|---|---|---|---|---|
| Age at ALF, y, sHRa [95% CI] | ||||||
| <2 | 1.00 [0.51–1.98] | 0.98 | 0.76 [0.2–2.91] | 0.69 | 1.13 [0.49–2.6] | 0.78 |
| 2 to <5 | 0.81 [0.39–1.68] | 0.57 | 0.37 [0.08–1.84] | 0.22 | 0.99 [0.42–2.32] | 0.98 |
| 5 to <10 | 0.58 [0.29–1.2] | 0.15 | 0.23 [0.05–1.12] | 0.07 | 0.79 [0.34–1.84] | 0.58 |
| ≥10 | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Sex, sHRa [95% CI] | ||||||
| Female | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Male | 1.38 [0.84–2.26] | 0.2 | 1 [0.38–2.59] | 1 | 1.4 [0.77–2.54] | 0.27 |
| Race/ethnicity, sHRa [95% CI] | ||||||
| White | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Black | 0.87 [0.3–2.6] | 0.81 | 0.37 [0.04–3.38] | 0.38 | 2.1 [0.58–7.62] | 0.26 |
| Hispanic | 1.34 [0.69–2.58] | 0.39 | 0.7 [0.21–2.35] | 0.56 | 1.8 [0.77–4.23] | 0.18 |
| Asian/Pacific Islander | 1.82 [0.72–4.59] | 0.2 | 2.39 [0.57–9.93] | 0.23 | 1.54 [0.42–5.65] | 0.51 |
| Native American/Alaska Native/Other | 1.41 [0.18–0.95] | 0.75 | NA | NA | 4.65 [0.52–41.26] | 0.17 |
| Etiology, sHRa [95% CI] | ||||||
| Indeterminate | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |||
| Viral | 0.13 [0.02–0.95] | 0.04 | NA | NA | 0.14 [0.02–1.09] | 0.06 |
| Acetaminophen toxicity | 0.59 [0.08–4.42] | 0.61 | NA | NA | 1.79 [0.23–13.89] | 0.58 |
| Other immune-related | 1.22 [0.28–5.25] | 0.79 | NA | NA | 1.19 [0.27–5.3] | 0.82 |
| Metabolic/genetic | 3.05 [1.06–8.73] | 0.04 | 2.68 [0.34–21.27] | 0.36 | 6.4 [1.84–22.26] | 0.003 |
| Other | 0.13 [0.02–0.94] | 0.04 | 0.54 [0.07–4.12] | 0.55 | NA | NA |
Italics denotes statistically significance (p<0.05).
Subdistribution hazard ratio (sHR). Analysis using Cox proportional hazards models demonstrates that metabolic/genetic causes of PALF were associated with a higher chance of receiving LT, especially in patients treated at transplant centers. Other causes of PALF were associated with a lower chance of receiving LT in aggregate.
Abbreviations: ALF, acute liver failure; NA, not available; LT, liver transplant; PALF, pediatric acute liver failure.
DISCUSSION
This large, population-based study examining PALF over a 28-year period establishes that the incidence of PALF is 16.75 per 100,000 live births in a representative sample of the US population. This is comparable to initial estimates of 500–600 cases yearly in the United States.6 Overall, 95.2% of patients lived in an urban environment, which is congruent with US Census data from 2020, which report 94.2% of California’s overall population resides in urban areas.25 Using this Californian population, age and disease etiology appear to influence PALF outcomes. Overall, children <10 years of age at presentation are more likely to receive emergent LT, resulting in children 5–10 years old experiencing lower mortality, especially following MELD/PELD score use in organ allocation in 2002 (Tables 2 and 3). Primary admission to transplant centers increased the likelihood of LT, especially in patients with metabolic and genetic diseases presenting with PALF. This study supports expedient referral and evaluation of patients with PALF at specialized transplant centers, given their increased risk of mortality and need for emergent LT.11,26,27
By examining a large population with longitudinal data, our study demonstrates that disease recurrence and liver-specific sequelae following the initial presenting illness in PALF are rare. Recurrent acute liver failure is defined as 2 episodes of acute liver failure interspersed by “complete recovery.”28 Cases of recurrent acute liver failure are anecdotally rare and described largely through case reports, with causes linked to genetic defects related to vesicular trafficking such as NBAS, LARS1, RINT1, and SCYL1.28,29,30,31,32 We examined all-cause mortality after 4 months and in the years after diagnosis. This type of analysis was possible due to comprehensive and high-level population-based registry data that captures all hospitals in California, as most centers discharge patients following SNL with limited long-term follow-up. We observed that LT and mortality risk largely peaked within 4 months of diagnosis (Figure 2), and late deaths were attributed to either hematologic malignancy or cardiac causes rather than hepatic causes. It is possible that some of these patients harbored pre-existing hematologic or cardiac conditions that made them otherwise ineligible for transplant, thus contributing to the mortality rate. Despite the predominance of indeterminate patients (55.4%) in this cohort, we did not observe a significant proportion of recurrent acute liver failure cases. Similarly, anecdotal evidence postulates that seasonality may be linked to PALF incidence, reflective of potential etiology, with a trend toward increased acetaminophen and indeterminate cases during the cooler months in the original North American study.6 This is perhaps with respect to potential unidentified viral infections within the indeterminate population, although our results fail to demonstrate a relationship between time of year and PALF development (p=0.54).
We observed an overall mortality of 18.7%, comparable to North American academic centers (PALFSG, 17.1%)2 and population-wide estimates in Germany (35/137, 25.5%).11 The number of acetaminophen toxicity cases in this cohort is lower than in comparable studies (6.1% compared to 14%), as discussed later in our limitations.6,33 Patients with acetaminophen toxicity have a discrete and identifiable liver injury event with available targeted therapy (n-acetylcysteine) and often experience better outcomes with lower risk of mortality and a lower need for LT, which likely reflects the decreased transplant likelihood seen within our cohort for patients over the age of 10 (as compared to younger patients who present with indeterminate diagnoses) (Table 2).6 With respect to age, we observed higher mortality rates for younger children with PALF, which is consistent with prior registry and multicenter case studies.3,6,33 Procedural and biological concerns play into this, especially for small children <16 kg who often require technical variant allografts (split or living donor), thereby limiting potential organ availability and increasing waitlist times.3,11,34 Interestingly, we observed a higher frequency of mortality in patients initially admitted to transplant centers versus nontransplant centers (26.5% vs. 14.6%), although the duration between PALF diagnosis and death was also longer (31 vs. 7 d) (Table 5). This is likely because transplant centers more often receive sicker patients, with younger median ages (10 vs. 12 y old) and more indeterminate cases (Supplemental Table S3, http://links.lww.com/HC9/C355).
In contrast, our cohort had an 8.3% transplant rate, much lower than previous estimates provided by the PALFSG at 32.1%2 and population-wide in Germany at 22%.11 We attribute this phenomenon partially to our broad capture of patients with PALF, including those that present at community and nontransplant sites over the course of the study timeline of 1990–2018, as compared to transplant-focused sites in the PALFSG. Comparable population-wide studies, which do capture community sites, such as the one in Germany,11 were conducted between 2016 and 2018 and likely reflect a more contemporary practice of LT in this patient population, which is also highlighted through policy change in our data set. We observed an increase in the likelihood of LT in younger children following the 2002 implementation of MELD/PELD-based organ allocation. Although policy changes have somewhat standardized clinical evaluation,7 subjective signs of PALF, such as encephalopathy, can still be difficult to grade in younger patients, leading to a reliance on aggressive definitive treatment through LT.1,7 In our population, younger children are associated with LT (Table 3). We observe that younger children were more likely to be treated at transplant centers, which also contributes to a higher rate of LT in this population (Supplemental Table S3, http://links.lww.com/HC9/C355). The referral volume contributes to improved clinical expertise at these specialized LT centers.4,35 This supports our findings that increased proximity to transplant centers, as urban dwellers, will also increase the likelihood for LT (Supplemental Table S3, http://links.lww.com/HC9/C355).36
A major strength of this study is the large and highly diverse PALF population, with longitudinal data to elucidate long-term outcomes. These data include all hospitals in California across a patient’s lifetime of residence in the state, rather than tracking isolated admissions or only patients enrolled in active clinical studies. This improves the adaptation of the data to the general American populace, as California represents ~10% of the total US population. This study is limited by its retrospective nature and its reliance on registry data and diagnostic codes, without granular patient-level data for additional clinical or laboratory findings. We were not able to link this data set to others, such as the US transplant registry, and therefore, waitlist mortality and transplant trends could not be studied among patients listed for LT. The use of diagnostic codes limited our ability to appropriately include contemporary disease subtypes (such as “activated T-cell hepatitis”) or DILI outside of acetaminophen, exclude acute-on-chronic liver disease, or patients who do not meet the PALFSG criteria. This may also explain our cohort’s low proportion of acetaminophen-related cases, as those without a concomitant PALF diagnosis were not included in our study. In addition, this study population terminates in 2018 and does not capture the recent spike in worldwide PALF cases coincident with the COVID-19 pandemic, and may not reflect the most up-to-date trends in disease development.
In conclusion, this large, diverse US population-based study identified risk factors associated with mortality and LT in children with PALF. We have established the incidence of PALF and did not observe seasonal trends in admission. While age was not a risk factor for death, younger age was associated with a higher chance of LT. Admission to a transplant center was associated with LT and mortality outcomes, especially for patients with drug-induced, viral, and metabolic/genetic causes. Patients with PALF who experience SNL appear to recover fully following the acute disease with no late liver-related mortality. Future studies are necessary to identify factors that predict disease outcomes and to develop triage tools that facilitate expedient transfer to specialized transplant centers. In doing so, we hope to maximize available resources and treatment outcomes in those who experience PALF.
Supplementary Material
Acknowledgments
FUNDING INFORMATION
Jason T.C. Lee is supported by a CIRM Clinical Research Fellowship in Stem Cell Biology and Regenerative Medicine. Johanna Ascher Bartlett was supported by USC Stem Cell’s Broad Clinical Research Fellows Program, the OneLegacy Foundation Fellowship Training Grant, and a Saban Research Institute Fellowship Award. Juliet Emamaullee was supported by a K08 from the National Cancer Institute (K08CA245220), American Society of Transplant Surgeons Faculty Development Grant, American Society for the Study of Liver Diseases Clinical, Translational, and Outcomes Research Award, and Liver Scholar Award from the Gilead Research Foundation.
CONFLICTS OF INTEREST
The authors have no conflicts to report.
Footnotes
Abbreviations: HCAI, California Department of Health Care Access and Information; ICD, International Classification of Diseases; LT, liver transplant; NIDDK, National Institute of Diabetes and Digestive and Kidney Diseases; PALF, pediatric acute liver failure; PALFSG, Pediatric Acute Liver Failure Study Group; PELD, pediatric end-stage liver disease; SRTR, Scientific Registry of Transplant Recipients; sHR, subdistribution hazard ratio; SNL, survival with native liver; UNOS, United Network for Organ Sharing.
Supplemental Digital Content is available for this article. Direct URL citations are provided in the HTML and PDF versions of this article on the journal’s website, www.hepcommjournal.com.
Contributor Information
Jason T.C. Lee, Email: jason.lee@med.usc.edu.
Arianna Barbetta, Email: arianna.barbetta@med.usc.edu.
Johanna Ascher Bartlett, Email: johanna.ascherbartlett@utsouthwestern.edu.
Sarah Bangerth, Email: sarah.bangerth@gmail.com.
Shannon Jordan, Email: sjordan9@usc.edu.
Mary Lo, Email: Mary.Lo@med.usc.edu.
Sarah Barhouma, Email: sarah.n.barhouma@kp.org.
Laura Houshmand, Email: laurahoushmand@gmail.com.
Linda Sher, Email: linda.sher@med.usc.edu.
Rohit Kohli, Email: rokohli@chla.usc.edu.
Kambiz Etesami, Email: Kambiz.Etesami@med.usc.edu.
Juliet Emamaullee, Email: juliet_emamaullee@urmc.rochester.edu.
REFERENCES
- 1.Ng VL, Li R, Loomes KM, Leonis MA, Rudnick DA, Belle SH, et al. Outcomes of children with and without hepatic encephalopathy from the Pediatric Acute Liver Failure (PALF) Study Group. J Pediatr Gastroenterol Nutr. 2016;63:357–364. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Squires JE, Rudnick DA, Hardison RM, Horslen S, Ng VL, Alonso EM, et al. Liver transplant listing in pediatric acute liver failure: Practices and participant characteristics. Hepatol Baltim Md. 2018;68:2338–2347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Baliga P, Alvarez S, Lindblad A, Zeng L, Studies of Pediatric Liver Transplantation Research Group . Posttransplant survival in pediatric fulminant hepatic failure: The SPLIT experience. Liver Transplant. 2004;10:1364–1371. [DOI] [PubMed] [Google Scholar]
- 4.Bhatt H, Rao GS. Management of acute liver failure: A pediatric perspective. Curr Pediatr Rep. 2018;6:246–257. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lutfi R, Abulebda K, Nitu ME, Molleston JP, Bozic MA, Subbarao G. Intensive care management of pediatric acute liver failure. J Pediatr Gastroenterol Nutr. 2017;64:660–670. [DOI] [PubMed] [Google Scholar]
- 6.Squires RH, Shneider BL, Bucuvalas J, Alonso E, Sokol RJ, Narkewicz MR, et al. Acute liver failure in children: The first 348 patients in The Pediatric Acute Liver Failure Study Group. J Pediatr. 2006;148:652–658. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Squires JE, Alonso EM, Ibrahim SH, Kasper V, Kehar M, Martinez M, et al. North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition Position Paper on the diagnosis and management of pediatric acute liver failure. J Pediatr Gastroenterol Nutr. 2022;74:138–158. [DOI] [PubMed] [Google Scholar]
- 8.Chapin CA, Burn T, Meijome T, Loomes KM, Melin-Aldana H, Kreiger PA, et al. Indeterminate pediatric acute liver failure is uniquely characterized by a CD103+ CD8+ T-cell infiltrate. Hepatol Baltim Md. 2018;68:1087–1100. [DOI] [PubMed] [Google Scholar]
- 9.Berardi G, Tuckfield L, DelVecchio MT, Aronoff S. Differential diagnosis of acute liver failure in children: A systematic review. Pediatr Gastroenterol Hepatol Nutr. 2020;23:501–510. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Narkewicz MR, Horslen S, Hardison RM, Shneider BL, Rodriguez-Baez N, Alonso EM, et al. A learning collaborative approach increases specificity of diagnosis of acute liver failure in pediatric patients. Clin Gastroenterol Hepatol. 2018;16:1801–1810.e3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lenz D, Abdulaziz M, Peters B, Wagner M, Schlieben LD, Corman VM, et al. Paediatric acute liver failure: A prospective, nationwide, population-based surveillance study in Germany. J Pediatr Gastroenterol Nutr. 2025;81:653–662. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lu BR, Zhang S, Narkewicz MR, Belle SH, Squires RH, Sokol RJ. Evaluation of the liver injury unit scoring system to predict survival in a multinational study of pediatric acute liver failure. J Pediatr. 2013;162:1010–6.e1-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lee EJ, Kim JW, Moon JS, Kim YB, Oh SH, Kim KM, et al. Development of a prognostic score to predict mortality in patients with pediatric acute liver failure. J Pediatr Gastroenterol Nutr. 2020;70:777–782. [DOI] [PubMed] [Google Scholar]
- 14.Ascher-Bartlett JM, Bangerth S, Jordan S, Weaver C, Barhouma S, Etesami K, et al. CHALF Score: A novel tool to rapidly risk stratify children in need of liver transplant evaluation during acute liver failure. Transplantation. 2024;108:930–939. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Bangerth S, Vijayendra S, Ascher-Bartlett J, Etasami K, Kohli R, Emamaullee J. Strategic changes in organ allocation policy and outcomes in pediatric acute liver failure. JAMA Surg. 2025;160:102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ascher Bartlett J, Barhouma S, Bangerth S, Mejia V, Weaver C, Kohli R, et al. Finance, race, ethnicity, and spoken language impact clinical outcomes for children with acute liver failure. Pediatr Transplant. 2024;28:e14686. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.2026 HCAI. HCAI—Department of Health Care Access and Information. Accessed January 16, 2026. https://hcai.ca.gov/
- 18.HCAI 2024. Data documentation. Patient-level data quality and error tolerance levels. Accessed January 16, 2026. https://hcai.ca.gov/data/request-data/data-documentation/
- 19.2026Caucuses. California State Association of Counties. Accessed January 16, 2026. https://www.counties.org/about/caucuses/
- 20.Kawaguchi ES, Shen JI, Suchard MA, Li G. Scalable algorithms for large competing risks data. J Comput Graph Stat. 2021;30:685–693. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.OPTN 2022.Improving liver allocation: MELD, PELD, Status 1A, Status 1B. OPTN. (Notice of OPTN Policy and Guidance Changes). Accessed January 16, 2026. https://www.hrsa.gov/sites/default/files/hrsa/optn/policy-guid-change_impr-liv-alloc-meld-peld-sta-1a-sta-1b_liv.pdf
- 22.Deep A, Alexander EC, Brierley J, Damian M, Gupta A, McLin V, et al. Paediatric acute liver failure: A multidisciplinary perspective on when a critically ill child is unsuitable for liver transplantation. Lancet Child Adolesc Health. 2024;8:921–932. [DOI] [PubMed] [Google Scholar]
- 23.Squires RH, Ng V, Romero R, Ekong U, Hardikar W, Emre S, et al. Evaluation of the pediatric patient for liver transplantation: 2014 practice guideline by the American Association for the Study of Liver Diseases, American Society of Transplantation and the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition. Hepatology. 2014;60:362–398. [DOI] [PubMed] [Google Scholar]
- 24.OPTN/UNOS 2011.Summary of actions taken at the OPTN/UNOS Board of Directors Meeting—November 14-15, 2011 [IMPORTANT POLICY NOTICE]. UNOS. Accessed January 16, 2026. https://unos.org/wp-content/uploads/2011-NOV-BOD-PolicyNotice_Compiled.pdf
- 25.2026Bureau UC. Census.gov. Nation’s urban and rural populations shift following 2020 census. Accessed January 16, 2026. https://www.census.gov/newsroom/press-releases/2022/urban-rural-populations.html
- 26.Lenz D, Hørby Jørgensen M, Kelly D, Cardinale V, Geerts A, Gonçalves Costa I, et al. Etiology and outcome of adult and pediatric acute liver failure in Europe. J Pediatr Gastroenterol Nutr. 2023;77:115–120. [DOI] [PubMed] [Google Scholar]
- 27.Lal BB, Khanna R, Sood V, Alam S, Nagral A, Ravindranath A, et al. Diagnosis and management of pediatric acute liver failure: Consensus recommendations of the Indian Society of Pediatric Gastroenterology, Hepatology, and Nutrition (ISPGHAN). Hepatol Int. 2024;18:1343–1381. [DOI] [PubMed] [Google Scholar]
- 28.Sutton H, Saban OS, Cunningham J, Bandsma R, Sayed B, Ng V. Phenotype and long-term outcome in recurrent paediatric acute liver failure: Systematic review and individual participant data analysis. Liver Int. 2025;45:e70412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Staufner C, Haack TB, Köpke MG, Straub BK, Kölker S, Thiel C, et al. Recurrent acute liver failure due to NBAS deficiency: Phenotypic spectrum, disease mechanisms, and therapeutic concepts. J Inherit Metab Dis. 2016;39:3–16. [DOI] [PubMed] [Google Scholar]
- 30.Haack TB, Staufner C, Köpke MG, Straub BK, Kölker S, Thiel C, et al. Biallelic mutations in NBAS cause recurrent acute liver failure with onset in infancy. Am J Hum Genet. 2015;97:163–169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Li JQ, Qiu YL, Gong JY, Dou LM, Lu Y, Knisely AS, et al. Novel NBAS mutations and fever-related recurrent acute liver failure in Chinese children: A retrospective study. BMC Gastroenterol. 2017;17:77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Lenz D, Schlieben LD, Shimura M, Bianzano A, Smirnov D, Kopajtich R, et al. Genetic landscape of pediatric acute liver failure of indeterminate origin. Hepatology. 2024;79:1075–1087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Kulkarni S, Perez C, Pichardo C, Castillo L, Gagnon M, Beck-Sague C, et al. Use of Pediatric Health Information System database to study the trends in the incidence, management, etiology, and outcomes due to pediatric acute liver failure in the United States from 2008 to 2013. Pediatr Transplant. 2015;19:888–895. [DOI] [PubMed] [Google Scholar]
- 34.Firl DJ, Sasaki K, McVey J, Hupertz V, Radhakrishnan K, Fujiki M, et al. Improved survival following living donor liver transplantation for pediatric acute liver failure: Analysis of 20 years of US National Registry Data. Liver Transpl. 2019;25:1241–1250. [DOI] [PubMed] [Google Scholar]
- 35.Elserafy N, Thompson S, Dalkeith T, Stormon M, Thomas G, Shun A, et al. Liver transplantation in children with inborn errors of metabolism: 30 years experience in NSW, Australia. JIMD Rep. 2021;60:88–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Adler JT, Bababekov YJ, Markmann JF, Chang DC, Yeh H. Distance is associated with mortality on the waitlist in pediatric liver transplantation. Pediatr Transplant. 2017;21:e12842; [DOI] [PubMed] [Google Scholar]